Sources and Standards

Every article on Botany for the Curious ends with a numbered list of its sources, linked to the originals wherever they are online. This page sets out the standards behind those lists: which statements need a source, how scientific names and trees are checked, how diagrams are drawn, and how errors are put right.

What needs a source

A statement of fact needs a source, and a primary source is preferred to a summary of one. Some kinds of statement must carry a source: scientific names and classification, species numbers and distributions, evolutionary relationships, definitions of terms and diagnostic features, dates and historical events, the identity of a chemical compound and any reaction it takes part in, any claim to be the earliest or the first, and any claim that a trait serves a function.

Where only the abstract of a paper could be consulted, the source list says so, and the page claims no more than the abstract supports. Statements that can go out of date carry a year or an "as of" date.

Scientific names

Scientific names are checked against two published checklists: the World Checklist of Vascular Plants (WCVP), published by the Royal Botanic Gardens, Kew, and the World Flora Online (WFO) Plant List. Each accepted name links to its record in Plants of the World Online, the Kew database built on WCVP, so the current status of a name is one click away. A scientific name is set in italics the first time it appears, followed by its common name.

When WCVP and WFO disagree about a name or where it belongs, the page gives both and explains the difference. Neither checklist wins by default.

When sources disagree

No fixed ranking of sources settles a disagreement. Three things are weighed instead: the evidence (whether original research supports the claim, and whether its method suits the question), the coverage (a broad synthesis can be older on a particular point than a focused study), and the date of the last update. When sources of equal standing disagree, both are shown side by side with what each rests on.

Two kinds of tree

Every tree on the site is labelled, on the diagram itself, as one of two kinds. A classification diagram follows a ranked hierarchy of names, from order to family to genus, and shows how names are filed; it is not evidence of how lineages are related. A phylogenetic hypothesis is redrawn from named studies, such as the Angiosperm Phylogeny Group's APG IV system of 2016, and from the Open Tree of Life synthesis, with each branch point traced to the research that supports it.

Unresolved and disputed relationships are marked as such. A branch point in the Open Tree synthesis that rests only on classification data, with no phylogenetic study behind it, is drawn as unresolved. When a newer study disagrees with the synthesis, both are shown with their evidence, and the synthesis does not override the newer work by default.

Trees are redrawn from the topology in the cited studies, not traced from published figures, so anyone can redraw them from the same data. Each tree is followed by a line headed How this tree was drawn, giving the kind of tree, the data source, the supporting studies, the date of the data and how the figure was drawn.

Classifications shift as new studies appear, so tree pages are rechecked every 12 months. A tree whose groups have moved is redrawn with a new data date, and the change is logged as a correction.

Diagrams of plant form

Before a diagram of plant form is drawn, every label and diagnostic feature on it is matched to an entry in a botanical glossary or a flora (a work that lists and describes the plants of a country or region). A label that finds no match is deleted, not drawn from guesswork. Leaf outlines and other shapes are constructed from the definition of each term rather than traced from a published plate.

Definitions are written in the guide's own words, never copied from copyrighted glossaries. The glossaries and floras behind a page are named in that page's source list. Where two species look alike, the page does not tell them apart by eye: identifications follow the labels and descriptions in the sources, and the page says which source each one comes from.

Compounds and processing

Each chemical compound is pinned to its PubChem compound identifier (CID), recorded for the compound in Wikidata, and its structure drawing follows the structure recorded with that identifier. A CID proves only which compound is meant.

How a compound forms or changes while tea leaves or coffee seeds are processed is a separate claim. It is cited to food-chemistry or plant-physiology research, and where no such research has been found, the page stops at naming the compound.

Garden history

Garden pages work from public-domain primary sources: period plans, engravings and documents, and records held by archives and libraries. Axes and sightlines are drawn over a public-domain base plan, and the base plan is credited separately. Where no plan can be reused commercially, the page draws an independent schematic of the axes and zones and names the documents it follows; nothing is traced from a protected plan.

Garden pages cover history only. Current opening times, prices and restoration work are outside their scope.

Legends and evidence levels

Tea and coffee come with origin stories, and the stories are labelled as legends rather than told as history. Each one is given with the earliest document that can be traced and one of three evidence levels. A legend is a story, usually an origin story, that no document from the time of its events supports; a later account reports events in writing after they happened; a contemporary record was written at the time. A legend often survives only in a later account, so the story and the document that carries it are labelled separately. The evidence level is checked again whenever the story is cited.

Common is not all

Patterns in plants have exceptions, and pages keep the two apart. A pattern seen in many cases is written as common or usual, with a source, never as a law. A claim that a trait serves a purpose needs evolutionary or functional research behind it, and the page leaves room for other explanations: developmental constraint, historical accident and selection by breeders. Changes people make to a single plant or a picked leaf, such as pruning or processing, are described as handling, not as the result of selection.

Calculations and models

Some statements can be checked by calculation, and those are recalculated: the branching order of a tree from its cited data, geometric figures such as Fibonacci counts and the golden angle of about 137.5°, and the axes and proportions of a garden laid over its base plan. Where a figure is the guide's own calculation, the data line beneath it says "computed" with the date.

A recalculation confirms the arithmetic and the drawing. Whether the input is right depends on the source it came from. When a simple model reproduces a pattern seen in plants, the page says that the model draws the pattern and does not show how the plant makes it.

No invented numbers

Pages do not describe what groups of people think or do without data. A figure about gardeners, students or tea drinkers appears only with a named source, its sample size and its period. Statements about plant awareness disparity cite named studies and report their conditions and results.

Corrections

Errors are corrected in the open. Each correction is listed at the foot of the page it affects and in the Revisions and Corrections log, with the date, what changed and the source, and the revision date in the page's Cite this page box moves with it. Errors can be reported to contact@botanyforthecurious.com; the Contact page says what makes a report easy to act on.